Our research program has two main components. Firstly, to develop and test models of cognitive processes via study of the way these processes breakdown following brain injury or disease. Secondly, to evaluate the cognitive status of psychiatric patients in order to test hypotheses concerning possible neuroanatomic correlates of these disorders. Our studies have utilized priming paradigms that place minimal demands on attentional and retrieval processes to assess the integrity of object recognition, naming, visuospatial processes and memory in patients with Alzheimer's disease, We have obtained evidence in support of a model that posits that posterior cortical pathology results in a selective degradation of previously acquired knowledge, rather than an inability to retrieve information from intact knowledge stores. These degraded knowledge representations are, in turn, proposed to be responsible for impaired naming and to substantially contribute to poor memory in patients with Alzheimer's disease. Our studies of individuals infected with the Human Immunodeficiency Virus (HIV) have found slowed response times, impaired motor-skill learning and other subtle deficits consistent with involvement of subcortical regions of the brain in a subgroup of subjects. The significant relationship between task performance and concentrations of a potent neurotoxin, quinolinic acid, in the cerebral spinal fluid of the HIV+ subjects, suggested that the deficits may be primarily due to infection of the CNS. In contrast, adults with Obsessive Compulsive Disorders (OCD) performed normally on similar tasks suggesting that the types of deficits seen in patients with structural lesions of the basal ganglia and other subcortical regions may provide a good model for understanding HIV-related cognitive dysfunction, but not for adult OCD.